Report correct file sizes in inspect

bio2zarr/core.py

@@ -3,6 +3,8 @@
 import dataclasses
 import logging
 import multiprocessing
+import os
+import os.path
 import threading
 import time
 
@@ -45,6 +47,22 @@ def chunk_aligned_slices(z, n, max_chunks=None):
     return slices
 
 
+def du(path):
+    """
+    Return the total bytes stored at this path.
+    """
+    total = os.path.getsize(path)
+    # pathlib walk method doesn't exist until 3.12 :(
+    for root, dirs, files in os.walk(path):
+        for lst in [dirs, files]:
+            for name in lst:
+                fullname = os.path.join(root, name)
+                size = os.path.getsize(fullname)
+                total += size
+    logger.debug(f"du({path}) = {total}")
+    return total
+
+
 class SynchronousExecutor(cf.Executor):
     def submit(self, fn, /, *args, **kwargs):
         future = cf.Future()

bio2zarr/vcf.py

@@ -6,6 +6,7 @@
 import logging
 import math
 import os
+import os.path
 import pathlib
 import pickle
 import shutil
@@ -1509,14 +1510,12 @@ class VcfZarr:
     def __init__(self, path):
         if not (path / ".zmetadata").exists():
             raise ValueError("Not in VcfZarr format")  # NEEDS TEST
+        self.path = path
         self.root = zarr.open(path, mode="r")
 
-    def __repr__(self):
-        return repr(self.root)  # NEEDS TEST
-
     def summary_table(self):
         data = []
-        arrays = [(a.nbytes_stored, a) for _, a in self.root.arrays()]
+        arrays = [(core.du(self.path / a.basename), a) for _, a in self.root.arrays()]
         arrays.sort(key=lambda x: x[0])
         for stored, array in reversed(arrays):
             d = {

tests/test_core.py

@@ -179,3 +179,19 @@ def test_5_chunk_1(self, n, expected):
         z = zarr.array(np.arange(5), chunks=1, dtype=int)
         result = core.chunk_aligned_slices(z, n)
         assert result == expected
+
+
+@pytest.mark.parametrize(
+    ("path", "expected"),
+    [
+        # NOTE: this data was generated using du -sb on a Linux system.
+        # It *might* work in CI, but it may well not either, as it's
+        # probably dependent on a whole bunch of things. Expect to fail
+        # at some point.
+        ("tests/data", 4630726),
+        ("tests/data/vcf", 4618589),
+        ("tests/data/vcf/sample.vcf.gz", 1089),
+    ],
+)
+def test_du(path, expected):
+    assert core.du(path) == expected